Wednesday, April 6, 2005
I'll Never Forget That Face:
The Science and Law of the Double-Blind Sequential Lineup
By David Feige
Eyewitness identification is the dirty little secret of the criminal justice system. It comes as no surprise that mistaken identification is a leading cause of unjust incarceration. What is a surprise, is that simple changes in the way identification procedures are conducted can have an enormous impact on how many innocent people are wrongly identified. Sadly, few people in the legal profession have been paying attention to the burgeoning research in this area.
On November 18th, of 2001, a Judge in Brooklyn Supreme Court ordered the New York City Police Department to utilize a double-blind sequential lineup. The decision capped nearly a year of litigation over the issue-fought primarily in the Bronx. The decision in Matter of Rahim Thomas, and an earlier decision denying the same motion in People v. Leo Franco placed the judiciary squarely in the middle of the new battle over the way lineups are conducted.
The law in this area is desperately behind the times. In the last 15 years, a large body of social science research has begun to unravel some of the mysteries of eyewitness identification. The findings are frightening. In a recent study of DNA exonerations, 53 of the 63 innocent people wrongly convicted were convicted based on faulty eyewitness identifications. Other studies, some involving hundreds of erroneous convictions consistently rank eyewitness identification as the leading cause of innocent people going to prison and even death row.
These miscarriages of justice can be prevented. It should be the business of every court to insure that the miscarriages of justice are minimized. It can be done. Indeed, it has been done. Seeking an order for a double-blind sequential lineup affords a court the opportunity to employ a procedure shown to decrease the chances of an innocent person being wrongly identified by 50 percent, while being just as effective in correctly identifying criminal suspects.
I. The Science
A. The Problems With Eyewitness Identification
"The vagaries of eyewitness identification are well-known; the annals of criminal law are rife with instances of mistaken identification." United States v. Wade, 388 U.S. 218, 228 (1967). Both archival studies and psychological research support this acknowledged truth - that eyewitness identifications, which are among the most common forms of evidence presented in criminal trials, are frequently wrong.
A number of explanations for erroneous eyewitness identifications have been postulated. For example, studies have shown that the experience of being a crime victim, especially when that crime involves violence, produces stress far beyond optimum levels for cognitive functioning, thereby reducing the potential accuracy of an eyewitness's identification.
Studies have also shown that certain pretrial identification procedures - e.g., leading questions, positive feedback from police after making the "correct" selection from a lineup or photo array, or repetitive viewing of the same suspect - can have a distortive effect on the act of retrieving memory.
B. The Superiority of Blind Sequential Presentation in Photospeads and Lineups
When compared to the traditional simultaneous lineup procedure, sequential lineups produce a significantly lower rate of mistaken identifications. In fact, critical tests have shown that sequential lineup procedures routinely decrease the potential for false identifications by 50 percent. In one of the first empirical studies on the sequential lineups, 243 undergraduate students witnessed staged thefts. Five minutes after the staged thefts, half of the witnesses were presented with a simultaneous photo array containing six persons. The other half were shown the six photographs sequentially. Half of the witnesses in each presentation condition viewed culprit-present photo arrays while the other half viewed culprit-absent photo arrays. The presentation style - simultaneous versus sequential - had a significant influence on witnesses' identification performances. In the culprit-absent presentation, only 17% of those witnesses viewing sequential arrays made a false identification, as compared with 43% of those witnesses viewing simultaneous arrays.
These initial findings have been replicated repeatedly by other empirical studies. Indeed, another study in a different lab showed that 39% of eyewitnesses viewing simultaneous six-person lineup identified an innocent person as the criminal, as opposed to 19% mistaken identification rate by those witnesses who viewed suspects sequentially. Yet another study found that among subjects shown culprit-absent photo arrays, false identifications were made by 20% of subjects who experienced simultaneous presentation and 5% of subjects who experienced sequential presentation. In a survey of studies, sequentially presented photo arrays successfully reduced false identifications in five different experiments, each aimed at demonstrating the ability of sequential presentation to reduce the singular and/or combined impact of typical lineup biases.
And if that weren't enough, another study in which the actual perpetrator was not present, found something even more frightening: when suspects were displayed in a simultaneous lineup, the false identification rate was 72.2%, whereas, when witnesses viewed photographs of suspects sequentially the rate of false identification decreased to 38.9%.
C. Why it Works -- The Relative Judgement Problem
The superiority of sequential presentation makes intuitive sense when one closely examines the process. A witness looking at a simultaneous lineup is likely, despite any cautions, to want to pick someone. And in so doing they are likely to pick the person who most closely resembles the actual perpetrator even if it is not the actual perpetrator. The judgement that is made is a relative one. But beyond being intuitive, the relative judgement problem has been exposed by scientific studies.
Sequential lineups, by contrast encourage witnesses to make absolute judgments (i.e., compare a single face in a lineup to their memory of the culprit's face) instead of comparative or relative judgments (i.e., decide which of several faces most resembles the memory trace. As Wells explains it, "although an eyewitness could reason that a given lineup member . . . was a relatively better match to the culprit than was a previously presented member . . . the witness could not be certain that a subsequent lineup member (yet to be viewed) would not prove to be an even better match to the culprit than the one being currently viewed."
The research reveals that many witnesses who correctly identify the culprit in a culprit-present lineup would simply identify another suspect when the culprit's photo is removed. Thus it is clear, albeit scary that a witnesses' use of relativistic judgment process results in increased rates of false identifications in culprit absent simultaneous lineups. By eliminating the 'relative judgement problem, A sequential lineup or photospread can enhance the reliability of eyewitness identification without resulting in fewer accurate identifications.
What is clear is that the legal community has been slow to react to the sea change in the science of eyewitness identification. What is striking is that with no additional effort, no added costs, and no additional inconvenience, the police, prosecutors, and judiciary could each put a stop to a practice that results in innocent citizens going to prison or death row for crimes they did not commit. And all it takes to avoid these miscarriages of justice is simple single small step in the right direction. It is a direction which has been well illuminated by the social scientists who have gone before us. A direction which leads to only one thing: the continued identification of the guilty, and a better way to avoid the conviction of the innocent.
II. The Litigation Strategy
Most systemic changes in the criminal justice system can be accomplished by any of the three branches of government, and the double-blind sequential lineup is no different. Thus the legislative, executive and judicial branches of government are all legitimate battle grounds in which to wage the fight for a change in identification procedures. It goes without saying that we should try to meet with the police chief, mayor or governor to educate them about the science of eyewitness identification and convince them to adopt the double blind sequential lineup/photospread as the local policy. Similarly, we should make efforts to reach the legislature either through direct lobbying or through the media. Procedural or legislative change, of course, has the ability to affect all of the identifications done in a jurisdiction, and accomplish the change in one fell swoop. But when legislative solutions dissolve, and policy changes fail to materialize, there is always, of course, litigation.
A. Finding Cases
By the time most clients get to arraignment, it is too late to have an impact on how an identification procedure is going to be conducted. In the vast majority of cases, we are relegated to an ex post facto challenge to the identification procedure through a Wade hearing. As a result, it is not easy to find cases in which an effective challenge can be brought. Once an identification procedure has been done, everything subsequent is likely to be tainted-the witness, after all has already seen the person they identified at the identification procedure itself.
As a consequence, it is critical to act prospectively, in an attempt to influence the way lineups are conducted before they happen. Because of the volume of cases they deal with, public defender offices are often in the best position to find test cases. There are two places to look: cases where a defendant is already in custody and is wanted in connection with a new case, or cases where an additional post arraignment identification procedure is sought. In the first case, counsel can intervene by Order to Show Cause, and in the second, by filing a motion in response to the formal request to compel a defendant to submit to a post-arraignment identification procedure.
B. Yes You Can!-The Problem of Judicial Activism
Interestingly, the battle over double blind sequential lineups has not been waged over the science. In fact in both of the fully litigated cases, the prosecution conceded the superiority of the blind, sequential methodology. In addition, both the Franco and Thomas decisions, albeit in dicta, endorsed the use of the blind sequential method.
Rather than assail the science, the prosecution in each case chose to argue that a court did not have the authority to grant the requested relief. This argument, based in large measure on a quote from Illinois v. Lafayette, 462 US 640 (1983) invokes the specter of judicial activism, and suggests fundamentally that the court does not have the power to modify its own order. Nothing could be further from the truth, and yet empowering judges to do justice is the most difficult tasks in this kind of litigation.
It is important to remember first, that courts do have the inherent power to insure that investigations are conducted in a fair and reliable manner. Contrary to what the government argues, Courts have a clearly defined and critically important role as facilitators in the criminal justice system. And beyond this, the judiciary has the inherent power to grant reject or modify proposed orders. In almost all of these cases, it will have been the prosecution that brought the initial order demanding that a defendant be placed in a lineup. Courts are not bound to meerly sign or decline. They can do more-and have.
In fact, over the years courts have modified lineup orders in creative ways perfectly analogous to what the defense will be seeking. In United States v. Tyler, in the Western District of Pennsylvania, the government moved for an order that the defendant participate in a line-up. The defendant moved for a "blank lineup." The district court granted the defendant the right to conduct the blank as well as the regular lineup, and additionally authorized the subpoenaing of "persons who look like Tyler" to appear in the blank lineup. In an effort to stop the blank lineup, the government sought a stay from the Third Circuit, which was denied. See 878 F.2d 75, 755-56 (3d Cir. 1989). And Tyler is not alone. Other courts have modified traditional lineup requests to include the conduct of a 'perpetrator absent' or blank lineup. In light of the blank lineup jurisprudence, courts should be hard pressed to find that they lack the power or discretion to order the prosecution or police to use the blind, sequential protocol.
Still, if experience is any guide, courts these days are famously hesitant to do anything that smacks of judicial activism. And so it is sad but true that sometimes we must remind judges of their essential function. The Cannon of Judicial Ethics concerns itself largely with issues of conflict of interest. Political affiliations and fiduciary obligations are exhaustively covered. Precious little time is spent on the methods or aims of judging. But what is said is said first -- sealing its place as both mandate and aspiration is this: "A judge should uphold the integrity and independence of the judiciary". Nothing in the code suggests excessive or unusual deference to prosecutors or police officers. The functionaries of the executive branch -- litigants like all others--are not to be shielded from the requirements of justice, nor do they exist outside its ambit. The gold shield does not shield its possessor from the long arm of the judiciary, much as the government might like to think it does.
C. There are Advantages to Bringing the Motion Even if you Loose.
Not all of the applications for a double-blind sequential lineup will be won. On the contrary, victory remains rare. However it is important to see the application for a blind sequential lineup in a broader context. Even when the application is denied, and even when the police refuse your direct request to do the lineup blind and sequentially, there are benefits to clients. Litigating a Wade type hearing is different when the police have opted for a simultaneous lineup despite your request. Similarly, failure to use the better technique can give rise to great cross examination at trial. The failure to do a sequential lineup allows you to focus on the error rates in the standard simultaneous lineup, the research supporting the superiority of the sequential lineup, and, in certain cases, expert testimony on the various methods of identification. In short, even the denial of the motion can be a significant tactical advantage at trial. Jurors should wonder why the police persist in an antiquated practice that regularly identifies innocent people as the perpetrators of crimes they didn't commit.
III. The Empire Strikes Back
Of course, the district attorneys offices around the country are not going to sit idly by while defense lawyers and judges move the criminal justice system forward. On the contrary. As a result of the burgeoning litigation in New York , and a few filings outside of New York, things are beginning to happen. Despite their professed opposition, several of the DA's offices in New York have already assembled panels and met with the experts in order to decide whether to 'voluntarily' adopt the blind sequential protocol. This, of course has not prevented them from strenuously arguing that defense lawyers and judges have no business telling them what to do. And while they continue to 'study the issue' while arguing that the rest of the criminal justice system should leave them alone, they are also beginning work together to assemble a challenge to the science. The lynchpin of this challenge is an article (not yet published) by Ebbe Ebbesen which suggests that the adoption of the sequential lineup may be premature. The Ebbesen study, even though it is not based on any independent research, is sure to be pounced on by Judges eager to deny sequential lineup motions. And if it is, one of the next battlegrounds may be to go back to litigating Frye/Daubert hearings on the questions raised by the science. It is important to remember that all of the researchers who have done tests and studies have come to the same conclusion--double blind sequential lineups reduce the number of innocent people wrongly identified. The United States Department of Justice itself, in its guide for law enforcement suggests that sequential lineups produce more reliable identifications than simultaneous lineups. And though the guide makes no recommendation as to what state and local police departments should do, there was a sufficient scientific consensus on the advantages of the sequential lineup to endorse if not recommend it.
While District Attorney's around the country sit by and do nothing, the police continue to use identification procedures that regularly lead to the conviction of the innocent. And while legislative or executive changes would be nice, thus far, outside of New Jersey, neither police nor prosecutors have bothered to listen to the mounting social scientific evidence surrounding lineups and photospreads. As a result, the defense community is left, once again, to fight for the kind of change that any fair minded person would voluntarily adopt. If Justice Barrett's decision in Franco can be understood as a cautionary tale, Justice Kreindler's decision, based on essentially the same pleadings, should be seen as a beacon of hope. It is always worth asking for double blind sequential lineups. The client you save may be your own.
David Feige is the Trial Chief for the Bronx Defenders and a commentator for WNYC and NPR. He also appears regularly on Court TV.